Broadcom is using VMware Explore 2026 to make its case that VMware Cloud Foundation can serve as the foundation for enterprise private AI, unveiling a wide-ranging set of infrastructure, model, data, developer, and security capabilities designed to move AI workloads into production.
The announcements center on VMware Private AI Cloud, with Broadcom positioning VCF as the underlying platform for running inference, agentic AI applications, and traditional enterprise workloads alongside the data they depend on.
For partners, the breadth of the portfolio could expand private AI engagements well beyond infrastructure deployment into application modernization, data services, cybersecurity, governance, and ongoing AI operations.
VMware AI Factory targets faster private AI deployments
VMware AI Factory will provide the software-defined foundation underneath VMware Private AI Cloud, automating infrastructure deployment and Day 2 operations from bare metal through model inference.
Broadcom said the technology is intended to simplify what can otherwise be a slow process of configuring compute, networking, storage, accelerators, and AI software before organizations can deploy their first model.
It also provides organizations with tools to manage GPU resources and AI token economics as workloads scale.
Broadcom is supporting an open hardware ecosystem around the platform, giving customers flexibility across CPUs, GPUs, accelerators, and infrastructure vendors.
VCF adds leading AI models and model-as-a-service capabilities
Broadcom also announced validated support on VCF for models from NVIDIA, Google, NEC, Alibaba Cloud, and Z.ai, including Nemotron 3, Gemma 4, cotomi, Qwen3.7-Max, and GLM-5.2.
Using vLLM, VCF can support a broader catalog of open-source models and allow enterprises to make those models available internally as a service while keeping sensitive workloads and data within the private cloud.
Broadcom’s Private Cloud Outlook 2026 found 56% of enterprises are already running or planning to run production AI inference in private clouds.
Tanzu connects AI agents to governed enterprise data
Broadcom is also extending the strategy into the application and data layers through VMware Tanzu Platform, which the company calls the agent platform for VMware Private AI Cloud.
New AI-ready data foundations are designed to enable agents to use structured and unstructured enterprise information without requiring businesses to move data outside their private environments.
Broadcom is adding a deny-by-default architecture, a developer harness for building agents, a curated marketplace for approved models, tools, and data products, and governance controls over how agents interact with those resources.
That creates a potentially significant services play for partners tasked with preparing customer data for AI, modernizing applications, developing agents, and putting governance around how those agents access enterprise systems.
Broadcom adds security controls for autonomous AI agents
Security is another major component of Broadcom’s private AI pitch.
The newly launched AgentMinder independently verifies agent identities and evaluates their actions against defined missions, context, permissions, and risk before allowing agents to access enterprise resources.
Broadcom is pairing those controls with new VMware vDefend and Avi Load Balancer capabilities. VDefend will discover components including MCP servers, LLMs, data stores, and tools, identify unauthorized “shadow AI” activity, and use AI to generate intrusion detection and prevention signatures.
Avi is being extended to restrict agent access to unauthorized tools, identify anomalous behavior, and help prevent the exfiltration of sensitive data.
The company also introduced TrueSource, a commercially supported portfolio of verifiably built open-source software spanning Spring, Java, Python, Node.js, hardened container images, and data technologies including PostgreSQL, RabbitMQ, MySQL, and Valkey.
VCF expansion broadens the private AI opportunity for partners
Taken together, the Explore announcements illustrate Broadcom’s broader strategy for VCF: providing the infrastructure where enterprise AI runs, Tanzu to connect AI applications and agents with enterprise data, and a security and governance layer that controls how those agents operate.
For channel partners, that could turn the private AI conversation from a discrete infrastructure refresh into a much larger opportunity spanning hardware, cloud infrastructure, data modernization, application development, security, and ongoing managed services.
Need a refresh on Broadcom’s acquisition of VMware and the subsequent changes to the company’s partner ecosystem and technology approach? Our guide covers everything you need to know.